Latest / Unfinished Business / If Clients Can Just Ask AI, What Are They Paying You For? | Rory Sheppard
Transcript
- Rory Sheppard: I do think there are gonna be people in the service industry that hung out a shingle, and said, I'm a whatever, right? hanging the shingle and having the answer is not enough. how do I frame the question? How do I challenge the answers that I'm getting back from AI? how do I process that in the service industry to help my clients in a better way than just giving them the first answer? the answers are not enough, Kimberly Griggs: Well guys, welcome back to Unfinished Business. this is our monthly segment in the middle of mastery where Rory Sheppard and I get to take some of the challenges that business owners are working through right now and try to make sense of them together. quite frankly, in the age that we're in with AI, customers all of a sudden have access to many of the same tools that we are using as providers. So the big question here, Rory, that we'll start to explore is if my customer can do more themselves, you know, what are they actually paying me for? So that's where I want to begin. Rory Sheppard: Yeah, thank you for the softball. Yeah. Kimberly Griggs: Yeah, right? This will be this will be an interesting topic of discussion. But let's just begin with this question. You know, how Rory Sheppard: Okay. Kimberly Griggs: how legitimate is that fear right now? Like what are you seeing from businesses that you're talking to and businesses that you're working with in how they're, you know, starting to approach AI in the topic of, you know, hiring their services versus what their customers are using AI for. Rory Sheppard: Yes, I think how big of a concern I I would start with AI is in almost every conversation, somewhere along with my clients, right? So you know, we're talking about AI, we're talking about staffing, we're talking about supply chain, there's some standards that we talk about on a pretty regular basis just because we keep thinking things are gonna get better and the you know, the workforce is gonna come back or The supply chain's gonna settle down, or we're gonna understand what we're supposed to do with AI. I would start with we still hear horror story after horror story of companies that have invested heavily in AI and not gotten a return. so what does investing in AI mean? And then what does that mean against service-based businesses? So You know, the big companies that are spending, literally millions of dollars, embracing AI, the numbers are still it's probably a little bit better, but the you know, ninety to ninety-five percent of those projects are not getting the return, any return or at least not the return that they expected. So we're still kind of in the infancy of how to use it, what to do with it, what choices should we be making. Now, if I go directly to your question of I'm a service based business that can use, any of the AI tools and ask the question, give me a marketing plan, give me a business plan, give me a fill in the blank, and they get one, right? That seems pretty complete, seems pretty real. And so I think the shift that we need to think about from a service based point of view is we're moving from an age of having the answer, And that being enough, right? So here's how you do a business plan or here's how you do a marketing plan. And so I'm the person with the answer to no, I'm the person not just with the answer, but with expertise to help you implement that plan. And so I think if you think about even a business plan, yes, you can generate a very pretty 40 page business plan. But what questions did you ask? What data did you give it? How did you validate the data? I'll give you a quick analogy and then we'll see if this I mean, I think this holds true, but One of my friends was a real estate broker. and so in the early two thousands, everybody was making a killing in the real estate business. basically zero percent down, kind of crazy financing. Everybody could be a real estate agent and make a ton of money. And one of the things that Alan told me early on during the kind of the financial crisis of two thousand eight, two thousand nine was we're about to get to a spot to where only the real real estate people can make a living. Because you've got to work hard, you got to have the experience, you got to have the knowledge. It's not just as easy as listing a property and selling it. he was spot on. there were so many people that were in the real estate business in 07 That weren't in the real estate business in two thousand and nine Kimberly Griggs: Yeah. Rory Sheppard: And so I do think there are gonna be people in the service industry that hung out a shingle, and said, I'm a whatever, That Now that just ha hanging the shingle and having the answer is not enough. It's gonna be how do I frame the question? How do I challenge the answers that I'm getting back from AI? how do I process that in the service industry to help my clients in a better way than just giving them the first answer? I don't know about you, I never trust the first answer. Right. I challenge it, say tweak this, hate that language, right? Whatever it is. I think the the first answer for me is Expertise is not enough or won't be enough as as we move into the future. But I do think if you've got the skills to do the work without AI, then you will actually still be able to have a premium value on the services that you deliver because it's not enough. we've talked about this before, but I can go listen to experts on podcasts or watch YouTube videos or read books. but then I've just got a bunch of voices in my head and I don't know which one I'm supposed to listen to. And Kimberly Griggs: Yes. Rory Sheppard: being the trusted voice, being the one that's got the knowledge, the experience, the the ability to help you walk through that and discern the right answer for you, not the generic answer that I got from Claude or Chat GPT. I mean, I think that's gonna be where it's Critical. Kimberly Griggs: Yeah. So when it comes to a service-based business providing that type of value, how how do businesses start to reposition the the value of value, you know, the value of expertise. Rory Sheppard: Yeah, So I'll go back to my friend Rob Rodriguez and his definition of value, right? Price is what you pay, value is what you get. And so I do think there is a a shift that's got to happen in the service-based industry of moving from pricing models that are based on hours worked, and really deliverables based pricing. because absolutely. the the research, analytics, you know, text generation piece of the work that used to take 20 hours. I can compress down to two to three hours for sure. so I can't bill you for 20 hours of analysis work when that's 30 to 45 minutes now, I'm gonna have to say the the deliverable that you get is this, not how much time it took me to generate it. I mean, there are some old school practitioners that everything's hours based, right? And so how do I make that shift? I think that's a real challenge for them, but something they're gonna have to do. And then I start talking about, you know, the value of the deliverable, not the value of the work, not the price of the work, but the value of the deliverable. And then my expertise maybe even showing, I mean, I haven't done this yet, but I mean, here's the generic chat GPT answer, right? Or the generic Claude answer. But then here's the work that I produced. I mean here's the so you get the difference between what you started with and what you ended with. I I I don't know if we've talked about this, but I think it's a you know college classrooms, right? I mean, do they ban the use of AI or do they allow the use of AI? I think, the the educators that are probably making the biggest difference right now are saying, Absolutely it's a tool you should use. But show me what you started with the generic AI version and then show me what you completed, right? I mean Show me how you used and interacted with it as opposed to saying it was all original work. so I think the shift is not trying to not use AI or saying I didn't use AI, but saying absolutely I used it. This is what I used AI for. This is the value I got from it. But then here's the additional pieces that I added because of my knowledge and experience. Kimberly Griggs: Right. And I think you make a good point and also Rory Sheppard: Mm-hmm. Kimberly Griggs: a relevant one because I think on August 11th, Claude announced that they are actually watermarking the text in not a way that you can see it on from the user perspective, but actually in Rory Sheppard: Right. Kimberly Griggs: the analysis or the predictability of the word, which I thought was really interesting because commentators that I've seen about that subject have started talking about exactly what you're saying is like the originality of the work. So sure, it essentially if you used AI to create the first version of anything, whether that's a document, a legal document, a business plan, a marketing plan, whatever, to scour the internet, scour examples, make make that relevant and think about things you might not have thought up initially in putting those things together. But still, what if I'm hearing you say correctly, it's like the contribution that you're making comes from the diagnosis. things that you've Rory Sheppard: Yeah. Kimberly Griggs: seen within that plan that have been an ultim you know an utter failure versus simpler tactics or different tactics or shifts that would improve the performance. Is that what I'm hearing? Like we're gonna as sellers of services, we're gonna have to learn to be able to communicate the difference. Rory Sheppard: Yeah. And the models are gonna get smarter, right? So I mean I Kimberly Griggs: Absolutely. Rory Sheppard: I did do a little bit of just I mean it's always intriguing to me, but if I just take the three major the big th the three biggest players, right? So if I take anthropic with with Claude, I take Open AI with Chat GPT, I take Google with Gemini, and say how many major minor releases that are substantial, not minor code fixes, but major minor releases have they done year to date. So they've between the three, they've done 23 releases on their tool sets since the start of 2026. So if they stay on that pace, you know, it will the number will be around 35 to 36 by the end of the year. So the pace of change in the tool sets is far outrunning what we can do from strategy planning, analysis, and dramatically better. I mean, a year and a half ago, even a year ago, we were talking about how to get more intelligent in the way we wrote our prompts for AI. if I look at the going back to what are those major releases from the big players, it's all around agentic AI. How do I start having agents do the work for me? How do I delegate full bodies of work and say, do this, then that, then this, then that, provide me the feedback. and so I think we also have to make a shift all of us, but especially in the service space, we need to get better at not doing the work, but kind of leading, guiding, directing the work. and it's a little bit of the shift that we've talked about over the last couple of sessions on how do I move from being the doer to being the leader. But now I've got to do it with a a set of chat bots or agents, AI agents that I'm doing the delegation to. And I'm gonna go back to Ken Blanchard's teaching on situational leadership. I'm gonna start with being very much a micromanager on what I'm delegating to that that agent and I'm checking their work all the time. Right. And then how do you grow that that knowledge and that experience over time? Right. But it should be my unique knowledge and expertise, not The generic off the internet experience. Kimberly Griggs: Right. So, you know, let's let's talk a little bit about what that transition looks like. I mean, if that's something that everyone should be practicing theoretically, but especially professional services, you know, what skills do do these providers need to be developing now? And, you know, how do we move higher up the chain rather than, you know, just competing on the production? Rory Sheppard: Yeah. I think that ship has sailed, right? We're not I mean AI can do the production. Kimberly Griggs: Yeah. Rory Sheppard: It's what makes I mean, in your space, what makes the difference between a good agency and a bad agency? It's the the knowledge, the skills, the experience of the leaders that are saying, I know they asked for this, but we need to deliver that, I'm reframing the problem. And the solution, because of my knowledge and experience. but I cannot tell you how important it is to check the work. So, I mean, we have to be way better at proofreading and checking the work of AI than ever before, I was working with a client just this past week and kudos to them. They had spent hours working with Claude, developing a set of skills, and then had it generate a whole bunch of work for them. it in you know, two or three hours generated what would have taken them normally, a week, maybe two to do. I mean, it was dramatic in the difference. But I'm reading through the work, right? And I'm going I pretty much know everything about your business, but here's an acronym. I have no idea what this is. what is this acronym? Kimberly Griggs: Yeah. Rory Sheppard: And they looked at it and they said, I have no idea what that acronym is. I mean, and so we had a, I mean, where it came from, who knows, right? But it was a whole bunch of documentation that throughout that single document, it had acronyms and phrases that didn't connect to anything in the business. Right. So Kimberly Griggs: Yeah. Rory Sheppard: not the horror stories we've heard of hallucination, but still I need to go validate that work. how bad would it be, in especially from a professional services point of view, to provide a client with a proposal or a body of work that had something in it didn't make sense? Kimberly Griggs: Mm-hmm. Well, and you know, to to that point too, I think language and and being able to communicate your expertise, part of the way that you showcase that is by being able to simplify. So, you Rory Sheppard: Mm-hmm. Kimberly Griggs: know, that that has happened to me with a couple of clients. And just as a provider working with small businesses and getting to know that business, you start to understand terminology and and verbiage that humans would really use to communicate that versus you know a machine and so Rory Sheppard: Yeah. Kimberly Griggs: really just you know clarifying what that actually means and so to your point it's like using it's a great starting point but actually I think it the opportunity is actually that it helps you improve it all the better because it helps you to really get very clear on Rory Sheppard: Yeah. Kimberly Griggs: you provide what the benefits are and what's in it for you know the the end user. Rory Sheppard: Yeah, and I It may just be me, but I mean, I can say fourth grade language, completely understandable, easy to I mean, use my voice, not I mean, don't use those words, right? I mean, that's crazy pretentious and I hate that. Don't do that and It it's good and remembers, you know, for one or two or three tries, right? But then I'm back to twelfth grade language and stuff that I wouldn't I mean, would never come out of my mouth. Right. I mean I'm like Kimberly Griggs: Yes. Rory Sheppard: I don't talk that way. I mean I I'm not gonna change, right? So I'm not writing proposals that way either. I'm not I mean Kimberly Griggs: That's really interesting too, kind of going back to the micromanagement piece. No matter how you're implementing this, yes, if you just have those things set on autopilot, that that has been consistent across the board for me. Is it is great for the once you've worked through some of those challenges the first two or three times, but then you Rory Sheppard: Right. Kimberly Griggs: go back to that same, you know. I'm sure everybody has projects set up now, but you go back to that same project. knowing that you're feeding the machine, you're feeding the knowledge, but yeah, it definitely does not pick that up and remember. So you definitely Rory Sheppard: Yeah. Kimberly Griggs: have to to work as if you are the boss sitting beside the employee. So you know, this brings to mind that I want you to remind us of that model. Cause I think that this is such an interesting framework. and forgive me for not remembering off the top of my head. But I think this is actually a critical skill for all of us in learning how to go from doer to to director. Rory Sheppard: Yep. So the the model is from Ken Blanchard. It's called Situational Leadership. and so the four kind of stages or phases of it are stage one is when they don't have any knowledge, skills, experience in the space. And so I am micromanaging them or managing them. But it I mean it is what we think of when we think about micromanagement. So it's I'm giving you the how to do it. I'm giving you the outcomes I expect. I'm checking on you, looking over your shoulder all the time, right? I'm because you don't know what to do. And so if I just gave you the work and said, go do this, you're gonna be super frustrated. And so I'm gonna manage you through that process. The next step is I'm not managing anymore. I'm collaborating with you. And so then I'm doing it in your just exactly what you said. I'm sitting beside you as the leader, right? Shoulder to shoulder, hip to hip, we're working on this together, right? And so I'm still sharing wisdom, knowledge, experience, but maybe you've got hands-on keyboard, you're doing more of it, but it's in collaboration. Third step is coaching. So then I'm on the sidelines watching, but giving you feedback, going, don't use those words, right? I mean, that's not my language, right? Back that off, change that, try again, right? And then finally it's into support, right? So you've got the skills, you've got the knowledge, you know how to do it. I'm here when you hit a a bump in the road or something that you did not expect to occur. I'm here for that support role. And so if I'm managing people, that's the process. I think if I'm managing a an agent in AI, whether that's I'm using co-work with Claude or I'm using Codex with Chat GPT. I'm doing that process where I'm teaching you along the way from management to collaboration to coaching to support. But that's the process. Kimberly Griggs: But in this realm, if we're to if we're managing agents as opposed to humans, how do you, you know, what processes should a business be thinking about and setting up to check the work, to go back and review the final output? Rory Sheppard: Yeah, so I think again when I'm in management or collaboration, I'm I'm reading it as if that first year intern is doing the work, right? and it's probably taking me longer than if I had written it myself. Right. I mean, so that's back to the I don't delegate because it's faster for me to do it myself. I think the same thing is true in this space for AI. The first few times that I'm doing pure management kind of delegation down to that agent in AI, it's gonna take me a little bit longer because I've got to do all the prep work. I'm gonna have it do it, but then I've got to check it all. And then correct it, send it back, work through it. I mean, it's gonna be slower than me just potentially physically doing it myself the first time. But if I teach the skill right and I I'm setting up projects or I'm creating skills in Claude to where that it's learning that and knowing what to do time after time after time. It gets better and faster as we go. but I still need to check, right? I've got some stuff I've been working on in the nonprofit space and we are hours into the process. And I had some stuff this morning. I went, where did I mean where did that language come from? you end up in these conversations like you're talking to a person, and it's like that's not how I mean there's nothing in anything we've done that uses that kind of language. So I mean, there's still a even at the coaching step, right? It's The the student thinks they've got it figured out, right? And then they start riffing on their own with with a whole new way to do it and that's not the standard. Kimberly Griggs: Right. So I mean, if we were to use that framework as a management principle, it's easy to see ourselves or at least for me, I I have a clear picture of the first three steps. Like you don't know anything at all, prompt. I am going to sit with you, do with you, you know, collaborate with you, prompt slash, you know, other expertise, other knowledge. pulling in resources off the internet, whatever you're using to to add to that or contribute to that process. Then coaching, meaning it's now starting to spit out some pretty decent output on the first go. But you know, there's some tweaks. I can see all of those in an in a in a workflow. But the fourth point is, you know, Rory Sheppard: Yeah. Kimberly Griggs: I guess my my question is is where's the productivity Rory Sheppard: When? Kimberly Griggs: gain? Yeah, at when, at what point are those first three steps complete enough for a business to know and not even just, you know, a few employees, but on a mass scale certainly, that it's time to integrate the agent. Rory Sheppard: Yeah, so I think I mean part of it is the complexity of the of the process. And then the other is the number of kind of unknown inputs that happen. Right. I mean, and so the more susceptible it is to unknown or unexpected inputs, I'm gonna have to hang out with it longer. you know, I think about all the work we did from a quality assurance point of view, even in the support model, there would be You know, random audits, random checks, doing validations. I mean, I think that's another huge question back into this. Not are you using AI but how do you use AI? What additional pieces did you put in? And then I'm still having to say, and I'm accountable for the output. I never relinquish the accountability step from me. and so what is the validation? How often am I doing the checks? are there flags that I can check on the output to go, are we where we need to be? and so I think that's the other shift is we need to start thinking more like quality assurance people that are validating processes that should be in control. But how do I know for sure? Right? Because Kimberly Griggs: Right. Rory Sheppard: that crazy acronym that just shows up in the text, right? and they, you know, gave me a stack of paper like this, you know, it and I just happened to flip to the one that had that crazy stuff in it, right? it doesn't eliminate any accountability or responsibility on our part to do some body of work. It's just at a different level that the work has to be done. Kimberly Griggs: Yeah, that's right. Rory Sheppard: And we're putting our seal of expertise on it to say the work product is good. And when you're doing all the steps yourself, you know that ninety nine percent of the time the work product is good. But there are still some times where you make a mistake. I mean Kimberly Griggs: Yeah. Yeah, absolutely. And then, you know, kind of again, going back to if you're a solopreneur professional service provider, this is an easier conversation to see yourself in because you're ultimately responsible and in control of the output. But when you start adding scalability to this is when it really gets messy because back to your earlier point in that so much waste, and we've talked to several professionals within the like managed service provider space where they're talking about burning credits left and right. So it's actually it's actually causing more waste and more Rory Sheppard: Yeah. Kimberly Griggs: decrease to the bottom line simply because they're experimenting with it. So while we're in this messy middle, what is a sound approach forward to start Getting familiar with its capabilities, moving through this model as you suggest. I mean, maybe we definitely hit the pause button on the agents, because again, if you're delivering services to someone else, the worst thing that could happen is to let something have access to, pertinent information that could be, scoured among the masses. So how do we start to build a transition plan or adopt a set of rules of how we're going to experiment and keep moving forward in the progression of this? Rory Sheppard: did we talk about waste and value added versus non-value added last time? Okay. Kimberly Griggs: We did, yes. Rory Sheppard: So I think for me, that's the place to start still. value added versus non-value added. So value added is something the customer would be willing to pay for. Non-value added would be Something that the customer would not pay for. And then in non value added, I've got non-value added but necessary, or non-value added and not necessary. So if it's non-value added and not necessary, I need to get rid of it immediately. If it's non value added but necessary, so payroll, I mean, customer's not willing to pay for that. I need to make it as lean and clean and efficient as I possibly can. if you're just starting out or you're trying to figure out how to like fully engage and use AI in your services firm or even non-service, I mean anywhere, I would say, what is that work that is non-value added but still necessary? Because that is where any dollar you spend there is expense that goes straight to the bottom line. it's non value added. Customer's not willing to pay for it, but I still have to do it. if I can Take those processes and do the automation steps, create agents to do that repetitive work that's non-value added but still necessary. that would be the first place that I would go. And then, over time, right? Then this is where the the high level of trust with the tool and the agents and your experience and expertise in creating them. Cutting it loose on something that's a value added piece. I would be very slow in that adoption rate. Because what you're selling in the professional services world in the future is the relationship and the and the trust between you and the person you're doing the work with. And that can get violated so quickly if I turn a process on too soon. And so I would be very quick to automate, systematize, eliminate those things that are non-value added and still necessary. but be slower on the ones that are the true value add that I'm delivering for the business. there's pieces that end up being value add, research, analysis, you know, pulling data, you know, gobs of data off the internet and bringing it back to me. I mean, all day long, right? Kimberly Griggs: Mm-hmm. Mm-hmm. Rory Sheppard: But I I am getting stricter and stricter with myself on now I need citations on all the facts and figures and data that you're giving me. Where'd you get that? I want to go validate as much of that as I possibly can, just to go, you said twenty three percent, but if I go look, is it still twenty three percent or is it a different number? Kimberly Griggs: Yeah. And I think it I I kind of like what you had mentioned before about, you know, really putting that to the test by asking the questions a different way. Because the other thing I think we haven't touched on, but that we know is is functioning that way, especially with the open open AI model, Chat GPT, is the bias that exists. And I'm very concerned about that model Rory Sheppard: Okay. Kimberly Griggs: in particular now that it's opened up to advertising. Not that we have, you know, how how it functions or whatnot, but anytime you start, you know, making money off of advertisers, Rory Sheppard: Yeah. Kimberly Griggs: it's pretty safe to assume that they're feeding that data back for a reason. So is it accurate? I mean, you know, and and that even brings a bigger point of just the validity of data, period. You know, how do how what sources are reliable when it's using the internet and all of these different data sources to compile? an answer. How do you how do you track the Rory Sheppard: Yeah, and I Kimberly Griggs: validity of it all? Rory Sheppard: I haven't been in that conversation recently. I was in a pretty good conversation, probably April, May time frame. you know, deep research. So if you've got something that is very dependent on lots of data, lots of sources, lots of analysis, I still think Gemini is the the best tool for that. it's tied to Google, right? And so I mean any and all data that's available is available to it at its fingertips. Think Claude, you know, Anthropic and Claude moved forward dramatically in the agentic space and in the business process space, you know, spring on and through the summer. I think Chat GPT's model with Codex has caught up a little bit, but part of this is where did you spend your money first, right? And how invested are you in that space? but guarantee, I mean, at the rate of change. Right. I mean one of those three major tools is releasing a significant release, whether major or minor, every ten days. Right. So I mean between the three of So once a month for every one of them. So if you're behind the day, I mean It's wait thirty days and there's a new release that's coming that changes something in that process. Yeah. What what is the what are the real facts, right? part of that for me is especially if I'm doing analysis for somebody, I'm gonna say there may be other sources, but the top three or four that I want you to go to are these, right? I mean I've already done the research to know where I want the trusted data to come from. Start with this and then we can navigate through the other data that you're pulling. But I wanna at least start the prompt with here's the data sources that I want you to trust first. 'Cause who knows where it'll go get data if I left to its own devices, right? It's just gonna go find could be an advertiser's data set, right? Whatever that is, but Kimberly Griggs: Are there have you found any resources helpful for better prompting, better, prompt language? I think that is kind Rory Sheppard: his name? Kimberly Griggs: of the challenge too. It's, you know, if particularly if you're looking for other methods that exist rather than just relying on your own experience, which is the is the beauty in in this tool. Rory Sheppard: Right. No absolutely. Kimberly Griggs: And it's almost like instant collaboration. But Rory Sheppard: Mm-hmm. Kimberly Griggs: in order to get a quality result that's not just data driven, something that you could find on the internet, I guess. Have you found any tools that you're particularly using or that you recommend to kind of better prompt engineer to get a to get a different or better result? Rory Sheppard: You know, I think I mean, I had some resources that I was kind of recommending a year ago. I've kind of walked away from them. because it's really your critical thinking skills that I mean, and this is not the best answer, I get it, but your critical thinking skills and once you read the response, right? What questions does that generate for you? That's the camp I'm hanging out in right now. I I probably ought to go back and look and see. I mean, I cannot think of his name His first name is Jay, but I cannot tell you his last name. he did some really good work early on, you know, prompting language, those kind of things. I'll see if I can get that to you so that we can put it in the show notes. but I like reading the results and then saying, Well, tell me a little bit more about this, or that that doesn't seem consistent with this other thing, right? I mean, let me challenge you with this. I just open question, right? So here's a tool and a technique that I started doing. is I have gone in and said, look over our last this last project or this last exchange. How how can I ask better questions? How can I challenge you to provide me information that is right the first time? Right. I'm giving it back to the tool and saying, tell me how I can interact with you in a different and better way. Kimberly Griggs: Yeah. Rory Sheppard: and it's been pretty good. Going back and forth and it'll give me a list of here's the data that you asked for and here's ten recommendations. I'll say recommendation one, confirmed. Recommendation two, open question. Here's a question for you know, recommendation two. Tell me about this or explain this or give me more detail here. Question three, confirmed. Question four, no, here's what I would rather you do, right? But using language that it's given me and saying, You're a better user and collaborate bet better with me when you do these things. But ask the tool. Right? I mean Kimberly Griggs: I I do like that Anthropic has has made that easier. So like when when it does have something within the the task or the the prompt that you've that you've talked about, it will ask you questions. But I I think in general I think something in my own experimenting that I've had to learn is to start thinking about questions ahead of just typing, here's what I want to do. I used to just not the same way that I would type in a search result, but maybe starting kind of in that line of thinking of this is what I want the outcome to be. And Rory Sheppard: Yep. Yes. Kimberly Griggs: then not really clarifying like, well, what role are you playing? You know, in what context what is the ultimate goal? Who is this for? Like, you know, it's filling in the details as the who, what, when, why, where and how, right? but starting with some of those questions on the front end has helped improve, you know, my output. Rory Sheppard: I'm not a I don't use voice to text. I don't record things on I mean I I don't do any of that, but I do have some folks that I've interacted with that are kind of in AI all day, every day. And one of their recommendations is just hit the microphone and start talking. Because you say things that you would never type. and you give more detail and there's more you expand more when you're just talking because it's fast and easy compared to typing. you know, we talk depending, right? I mean Texas is probably a little bit slower, but I mean, on average it's between two hundred and twenty and two hundred and for forty words a minute. typing, I mean We're not ninety words a minute typers anymore. So it's probably in the fifty to sixty kind of words per minute. So I can be four times as productive just hitting the microphone and talking as opposed to Kimberly Griggs: Yeah, I've had I've had several AI experts tell me the same thing. And it does make sense now that we're talking through it. It makes sense because you it's almost like your brain gets fatigued, or that's how I feel. Like I get fatigued with I'm typing, but I have built a habit in doing that. So kind of removing yourself and again, it's it's almost like an experiment. Everything in AI is an experiment. Rory Sheppard: Yes. Kimberly Griggs: testing different ways of working with the tool in order to see what's possible. Rory Sheppard: Yeah, so there was, I mean, there was a kind of a myth that was out there, it's probably been a year ago now, that when you speak to your AI tool with anger, it gives you better answers. Right. If you're if you're mean to your AI tool, it gives you better answers. but there was a a guy working with Gemini that went and did the analysis, and really what it is is When it gives you an answer that doesn't make sense or you get frustrated, you're more precise, you're more detail oriented, you get I mean, you expand on the what the criteria is because you're going, You idiot, I can't believe you did that. Right. I mean, but Kimberly Griggs: Right. Rory Sheppard: then you give it all the all the ammunition that you should have given it to start with. you give it the second or the third time when you're frustrated. And so I think You don't need to get upset and angry first, but I do think how much detail answering the who, what, when, where, why, how questions in context or upload the documents here's what I got from the client, or here's what I'm starting with, or here's client documents and here's the stuff for me. now I would, you know. Whichever tool is you've that's out there, I can say, you know, all this detail is is private to this project or private to this chat, not openly available. I'd set that on every project, every chat that you're doing is keep it local. So when I'm putting my proposal from my you know the RFP from a client or whatever it is, it's local to me, not available to everybody. Kimberly Griggs: Yes. Rory Sheppard: but then what am I putting in there to go with that? before I ask the first question. I mean, seeding it first as opposed to asking the generic question and then giving it a whole bunch of data. Give it data first. Kimberly Griggs: So making this just a little bit more practical and kind of wrapping it up because again, it just continues to unfold such at a rapid pace that I think talking about it is probably the best way forward in helping everyone through the transition. But if somebody is listening to this episode right now and they own, you know, in particular a professional service business, what quick things should they be evaluating right now? you had mentioned several, but let's kind of recap. Rory Sheppard: I would start with what is the expertise that you no- kidding have? Right. And so it's not just I hung a shingle and I've got the answers, the answers are not enough, what makes me different or what experience do I have what are the hard knocks that I've gone through, and have those top of mind. not just in my conversation with potential clients, but I need to be using that as a way to differentiate the output that I'm creating if I'm doing it with AI. Second one would be I think that the move from pricing per hour to pricing per deliverable or result. I think that's critical for us. And as I was saying that out loud, I've got a couple of clients that I need to go back and have that conversation with. because I think that's a that's a huge mindset shift for some of my clients, right? That have always been no we're completely transparent on the hours that we're working. And then we're billing you based on hours. I I mean, and there's still some industries where hourly rate is how all the contracts are negotiated. So I mean, you can't get away from it, but I think anytime I can get to deliverables pricing, I think that's gonna be a shift that we need to make. the third one, and you can probably throw in a couple more, but the third one for me would be I need to start. Going down the path of either standard processes that I've got prompts for or going into kind of the agents of going into Codex or Cowork and playing with those. And then the first work I'm going to give it are those things that are not value added but still necessary. So that those are costs that I'm eliminating from the start. those are costs, if I'm doing the per hour billing, those don't even show up there because the customer's not willing to pay for it. it's just the the back end grunt work that's part of overhead and it's getting covered. I need to be getting as efficient and effective as I can with that work. and that gives me the ability to learn the tools. initial Kimberly Griggs: business. Rory Sheppard: research and analysis, first drafts, right? I mean, things that you might give to a, early on in their career person or an intern or something like that. What's that first body of work? any of those that I know it's work that has to be done, but the customer wouldn't pay for it if I put it on a line item for the bill. They're gonna go Yeah, no, I'm not paying for that, right? I mean, that's expected. Kimberly Griggs: Yeah, it's an expectation. Rory Sheppard: Yeah, that's expected. So anything that fits that kind of definition ought to be automated. another conversation for another day, but the next crisis that's gonna happen is because I've given all that work to AI agents. I've got a a group of eighteen to twenty four year olds that don't ever get that grunt work as their first job. but ten years from now I'm gonna expect them to be middle management, but they don't have any of that grunt work experience. That's another crisis that is another day, but it's part of this that's real. Kimberly Griggs: Well, I mean, I that's a great place to end as far as suggestions because I think again, I the best thing that any of us can do moving forward is share information. So whether it be things like this or resources if you're a provider, I kinda think just in summary for me and what I'm taking in and also just trying to apply myself is to lean into where we may have been nervous to approach the conversation of how we are all implementing this in our our daily practices. Now I think Rory Sheppard: Yeah. Kimberly Griggs: the conversation is more about sharing, you know, sharing tips, how we are and being transparent about how we're using it. So that Rory Sheppard: Yeah. I mean Kimberly Griggs: you can start to create some guidelines and, you know, AI policies, which again, I think as things move forward in the originality of the bodies of work, I think having some of those policies in place is really going to matter. If we're if we're working to build a better foundation in this. Rory Sheppard: Yes. Yeah, I think it's not a question of did you use AI? I think that's the wrong question. It's how did you use AI? What authority or expertise, you know, what person with authority or expertise evaluated what AI gave us and corrected, updated, checked it anyway, right? And then what additional pieces did you do after you got that output? What creative, unique things did you do with it that I couldn't have done on my own? Kimberly Griggs: And then I guess I I'll I'll end with one final just, you know, share what what are you excited about where this goes, where this leads. Rory Sheppard: So the inverse of we're gonna have a bunch of kids that don't have the initial experience, so they can't be experts or middle management someday is that since 2007, we have been at a negative replacement rate on workers. So we're we're almost 20 years into this situation where baby boomers are at retirement age and retiring. And we're at a one point seven for every two people replacement rate, and have been for 20 years. And so 15% lower replacement than we had. And we have 10,000 baby boomers that are still reaching age sixty five every day in the United States. And so the exciting thing for me is that if you look at any other time when We use technology as a, fuel for economic growth. There's been a workforce that went with that. So industrial revolution coming out of World War II. I mean, we had a huge population of workers to take advantage of the technology. And today we're in a spot where we're completely and solely dependent on the technology. to provide economic growth for us. And so the exciting thing for me is that we actually have a way forward for economic growth that can be crazy powerful and do some of that repetitive work that none of us want to do. and so I think that's exciting. Instead of doom and gloom on, you know, we don't we're not gonna have the workers 10 or 15 years from now. It's we've got a technology that's stepping in to fill that space. and the the knowledge worker space is going to be more and more about how do I use those tools to be more productive, right? I mean, you think about the advent of the PC. We didn't I mean we didn't eliminate any workers, but we created a whole bunch more work to be done. Kimberly Griggs: Yeah. Well, we will definitely be linking all of the information for how to contact Rory if you're interested in having a deeper conversation or even just an AI readiness assessment of how your business might be looking forward. We definitely wanna make sure we link Rory's contact information in the show notes. And then I want to give you space to, while this has been a big discussion, there's a lot to think about, give you a space to to promote where we can digest all of this in our own thoughts and time. Rory Sheppard: Yeah, so we spent a lot of time talking about technology and how to maximize our productivity. I'm still gonna tell you I'm way more productive when I do stuff with a piece of paper and a pen, right? because it's all about getting the right anchors for me and the right rhythm in the way I live my life. And so My Anchor Points journal, you got some last week a a new version release on that. all the science says when I'm typing or using my phone and putting stuff in electronically, The neurons and the synapses, I'm firing 500, 600, 800 of those when I'm doing that activity. When I'm writing the same thing on a piece of paper with pencil or pen, it's more than 10,000 synapses that are going on in my brain. So my brain is significantly more active when I'm putting pen to paper. Going through the process of where is my gratitude today and what am I grateful for? Saying positive affirmations over myself about the day that I'm about to go into. No kidding, praying for others first, praying for me and my needs second, and then maybe just a couple of things that I want, and then what am I doing from a meditation and emotion point of view? Being very intentional about the rhythm and the process that I go through on a daily basis. it can significantly impact the way you just interact with family, friends, and coworkers on a daily basis. I love the rhythm, I love the routine. People that I have shared that tool with, I mean, they say the same thing. When I use it on a consistent basis and get into a rhythm, I can tell when I don't. Kimberly Griggs: It is a centering because you know the reality of everything that we've been talking about is fear. And I think that the the thing that I find the most value in, and I want to acknowledge you for taking the time to really think through the things that make a difference in us from a thought life perspective, is the fact that it is an anchor in the day and it is a way to process all the clutter. that we're intake. I mean, if you just Rory Sheppard: Yeah. Kimberly Griggs: think about what the the latest research says, but just the amount of information that we ingest and the fear that we have around that information, you it has to go someplace. Otherwise it just lives in us, right? And then you're Rory Sheppard: Yeah, it's yeah. Kimberly Griggs: not good to anybody. And you sure with clarity, you know, we start to find our way forward. or at least I have, you know, in in using that. Rory Sheppard: Yeah, and I we don't talk about it very often and it I mean it is it is one of my favorite parts of the things that I write down on a daily basis, but there's something called a focal practice that's in the journal and it's those non technology based, what is it that requires almost full attention from me? Or something bad's gonna happen, right? I mean, I need to be fully focused on it. almost all consuming that I can do, but also can be regenerative, right? for all of us it can be something different, but I mean, singing in the church choir or being part of a prayer team, or my wife quilts, right? Janet quilts. she can't quilt and do anything else without poking a finger, making a bad stitch, right? I mean, and so what is it that I'm going to do that can be all-consuming for me that gets rid of some of that clutter? we've got a huge garden, going out and being part of nature, right? What is it that I can do that recenters me on what's important as opposed to all the noise and all the things that I could get wrapped up in. another key component of the tool and tool chest to help make sense of the stuff that's going on all day, Kimberly Griggs: Yeah. Rory Sheppard: every day around us. Kimberly Griggs: Yeah. I love it. Well thank you for creating such a a beautiful and powerful tool. It has been a a a great reflection tool in my toolbox, Rory Sheppard: Yeah. Well thank you. Kimberly Griggs: Okay, well thanks guys for tuning in and we will see you next time.